Nguyễn
Hoàng Ninh, Phạm Văn Sơn, Nguyễn Thị Hải Vân,
Nguyễn Đức Hà, Nguyễn Thị Huyền,
Nguyễn Quốc Khánh, Nguyễn Thị Liên, Trần Văn
Trọng, Nguyễn Viết Tuân, Nguyễn Huy Dương
In recent years, advances in Earth-surface
measurement and remote sensing technologies have created a variety of digital
topographic data and satellite images. The development of data processing
technologies has made it easier to exploit the aforementioned sources in many
fields, especially effective in geohazards investigation and research,
including landslide study. In this paper, the results of 3D-relief analysis and
high-resolution image interpretation for landslide inventory mapping will be presented. A
pilot study area was carried out at Tam Chung commune, Muong Lat district,
Thanh Hoa province. As results, 4 old landslides were interpreted from
3D-relief and 222 new landslides were interpreted from Google Earth images. The
field surveys has approached 97 of them (equivalent to ~42.92% of the total interpreted ones). It
confirmed that 94/97 interpreted landslides had actually occurred. Therefore,
the accuracy of interpretation reaches ~96.91%, which have demonstrated the
effectiveness of the used method in identifying historic landslides. Although
there are still some shortcomings due to the availability of necessary data,
the results have shown the advantages of the used method when it is applied in
areas with a complete set of topographic and remote sensing data. The used
method can be further developed to monitor the change of the topographic
surface over time, thus, it can effectively serve the landslide mapping at
large scale.
Keywords: remote sensing, image interpretation, 3D-relief analysis,
landslide inventory mapping